This publication is a doctoral dissertation which investigates the reactions of different groups of native speakers to Dutch-accented speech. The initial goal is to establish two separate hierarchies of items for a pronunciation syllabus for Dutch learners of English: one for those who aspire to speak Received Pronunciation (RP) and interact with RP speakers, and an equivalent for General American (GA).

It is a report of two quantitative studies. The first is a written survey of Dutch teachers, students, and lecturers on their priorities for pronunciation teaching. The second study is a (much bigger) online survey involving native speakers from a variety of backgrounds who gave their reactions to recordings of sentences which contained errors characteristic of Dutch speakers of English. Interspersed in the account of the research is a well-articulated and (mostly) even-tempered discussion of the tenets of the English as a Lingua Franca movement.

The importance of this work lies in its gathering information on reactions to Dutch pronunciation not from a single monolithic group of native speakers but from groups of native speakers who have different varieties of English as their first language. Also interesting is the perspective of the fears that Dutch learners of English have about which groups of native speakers would be more strict, and which more lenient, towards Dutch-accentedness. According to the author, it turns out that contrary to the expectations of Dutch learners, American and Canadian native speakers of English are more strict in their evaluations of Dutch errors than British, Irish, and other groups.

Chapter 1 gives the background and literature review, establishes the objectives of the research, and describes the methodology used. There exists a large group of Dutch learners of English who aspire to communicate effectively with native speakers – and the research starts with the question of whether native-speaker severity judgements differ according to whether the judges are speakers of RP or GA.

Chapter 2 describes the two experiments. The first, smaller ‘Dutch’ experiment had the purpose of providing the input to the second, larger ‘Native Speaker’ experiment. The Dutch experiment consisted of paper and e-mail questionnaire of 261 university and college lecturers, students and pupils, and secondary school teachers who were asked to give severity ratings to typical Dutch pronunciation errors. The author then calculated an Error Severity Index for each item. And on the basis of this experiment, the author chose 26 errors in five categories as input to the Native Speaker experiment. The categories were:

• • phonemic – e.g. no distinction between the vowels in ‘bat’ and ‘bed’

• • realisational – e.g. using a uvular-r in red, lack of appropriate dark-l in full

• • distributional – e.g. pronouncing film as two syllables

• • stress – e.g. stressing the third syllable of advertisement

• • suprasegmental – e.g. insufficient weak forms, insufficient contracted forms, and ‘too little variation in intonation’, ‘stressing to in going to Wales’

The Native Speaker experiment was conducted by making two sets of recordings, one with the pronunciation errors contained in British RP accented recordings, and the other with the errors in GA accented recordings. These were recorded by two bilingual actors who could accurately mimic the Dutch mistakes.

Participants in the survey were asked an open question about their accent background, and they were then grouped, for the purposes of analysis, not into simply RP and GA categories, but into seven major accent groups which included British non-RP, Northern and Southern Irish, Canadian, and both GA and non-GA. A further broad category included Australian, New Zealand and South African accents. A total of 545 judges made 16895 judgements.

Participants selected which of the two sets of recordings they felt most competent to judge. Close to 60% opted for the RP version and 40% for the GA version of the experiment. They were also asked about their age, gender, and whether they believed themselves to be strict or lenient judges of pronunciation. For each of the tokens they were asked whether or not they detected an error, and if so to judge whether it fell into one of four categories from ‘very serious’ to ‘not important’. The output of the experiment gives a severity ranking to each of the tokens.

Chapter 3 is the longest chapter, at close to 140 pages, of which Section 3.5 is central to the research, as it describes the detailed results for each of the items which the informants were asked to judge. This is best used as a reference section as one reads the discussion of results. For both the RP and GA versions, word-stress and phonemic errors were judged most severe, and suprasegmental errors were rated least severe.

Chapter 4 looks at the issue of ‘accent similarity’, which is an attempt to measure the extent to which judgements were affected by the existence of a native-speaker non-standard pronunciation which matches the Dutch errors. For example, the pronunciation of film as two syllables has a counterpart in Irish varieties, and just over half of the Irish judges did not regard this as an error, but of the others who did, made comments which showed that they regarded it as a stigmatised form ‘common . . . among Southern Irish plebs’ (p. 181).

Chapter 5 is a discussion of the results, and the author once again nails his colours to the mast, writing of the importance of native-speaker attitudes to Dutch accent errors, because of ‘the sociolinguistic dominance of its native speakers’ (p. 286). Now that Indian, Chinese, and South American economies are becoming the biggest drivers of the world economy, I wonder if the author would hold to this point of view quite so strongly. Points made in this chapter include: the British, Irish and Antipodean judges report more errors than the North American judges, but diagnose fewer of them as serious; speakers of non-standard varieties make judgements based on the standard variety.

Chapter 6 contains the author's recommendations. They include (for learners using the RP model) teaching learners to avoid the /ʌ/ – /ɒ/ confusion ‘which is more important than warning them off “Americanisms” such as yod-deletion in new’ (p. 316); and (for those who want the GA model) avoiding the substituting of /θ ð/ by /td/. This latter advice allows the author to make another point: just because a pronunciation feature is attested in a variety of American, it does not mean that it is acceptable, as it may be a stigmatised usage.

The importance of the dissertation lies not so much in the two hierarchies of pronunciation errors (which do not differ by much) but in the discoveries that were made as the research proceeded. The author states (p. 311):

The dissertation is admirably well-written. It is absolutely clear what the author has done at every stage of the research, and he is honest about the wrinkles that appeared as his research progressed, and how they affected the course of his research and its final shape.

The value of this work for me is in the fact that it starts with very clear ideas and simple categories (RP, GA) and gently demonstrates that these categories, and the judgements that emanate from them, are of much greater complexity. It is an essential read for anyone whose research interests lie in the area of attitudes towards accents.